Improved Combined Inertial Control of Wind Turbine Based on CAE and DNN for Temporary Frequency Support

With the continuous and large-scale development of renewable energy, there is a prominent decrease in level inertia new power systems. This leads to weakening system’s capability provide support frequency regulation during disturbance events. The wind turbines (WT), as main representatives energy generation, should be more efficiently involved system dynamics. However, optimal difficult achieve through combined inertial control strategy because it greatly depends on parameters fluctuates different scenarios. To cope with quickly scenarios obtain results, this paper presents an improved intelligent WT based contractive autoencoder (CAE) deep neural network (DNN). method obtains for using particle swarm optimization (PSO) algorithm, then effectively extracts features from actual data CAE followed by building model predict online. verify test proposed method, applied IEEE 9-bus system. simulation results show that can faster computational time, good prediction accuracy, generalization capability.